Objective To examine trends in epilepsy admissions in children from 1981 to 2013.

Design Repeated cross-sectional, population-based study.

Setting Scotland.

Patients We identified admissions among children between 1981 and 2013 inclusive. Epilepsy admissions were identified from the Scottish national hospital discharge database by using relevant diagnostic codes. Primary epilepsy admissions (PEAs) were those with epilepsy as the primary discharge diagnosis, or convulsions as the primary diagnosis but with epilepsy as secondary diagnosis. All other epilepsy admissions were secondary epilepsy admissions (SEAs).

Main outcome measures Trends in annual epilepsy and non-epilepsy admission rates, as well as sociodemographic, clinical characteristics, length of stay and readmissions of epilepsy admissions.

Results 57 031 epilepsy and 3 863 809 non-epilepsy admissions were available for analysis. Overall, epilepsy and non-epilepsy admissions increased, with a greater increase in epilepsy admissions (interaction Χ² test statistic 252, p<0.00001). Elective epilepsy admissions, unlike elective non-epilepsy admissions, continually increased, but emergency epilepsy admissions increased until 2000 and showed only minor fluctuations thereafter. Increase in SEAs was more marked than PEAs (interaction Χ² test statistic 627, p<0.0001). 48% of epilepsy admissions were to children’s hospitals. No substantial trends were apparent in age, gender or deprivation distribution of epilepsy admissions. There was a clear trend towards shorter length of stay.

Conclusions Childhood epilepsy admissions are increasing, at a faster rate than non-epilepsy admissions, and have changed towards shorter, more elective admissions. Many will not be to children’s hospitals, and the primary reason will often not be because of epilepsy/convulsions. More, not less, epilepsy resources are needed.